Could Atoms Be Reordered to Enhance Electronic Devices?

Detail of a cover illustration created by U of A for the journal Advanced Electronic Materials.
Eric Pipkin

Detail of a cover illustration created by U of A for the journal Advanced Electronic Materials.

The optical properties of a thin layer of the semiconductor germanium-tin (GeSn) sandwiched between barriers of silicon-germanium-tin (SiGeSn), a structure known as a quantum well, have been studied with a focus on improving lasers and photodetectors. But what would happen to an electrical charge passing through the nanometer thick germanium-tin quantum well? 

Scientists predicted that the electrical charge would move relatively slower in the quantum well, because the barriers were made of a combination of silicon, germanium and tin. 

"We thought it would be worse, because we mixed things together. But we found the mobility is higher," said Shui-Qing "Fisher" Yu, professor of electrical engineering and computer science at the University of Arkansas and a lead investigator on the study. 

The surprising discovery could help advance both neuromorphic computing, which mimics the human brain, and quantum computers. The results may also help scientists understand the role played by the short-range ordering of atoms. If scientists can manipulate the ordering of those atoms, it could produce a leap forward in the performance and miniaturization of microelectronics. 

The research was conducted by a team from the U of A, the Department of Energy's Sandia National Laboratories, and Dartmouth College. 

"It's a very interdisciplinary team," Yu said. "You get this synergy from three groups working closely on a single problem." 

The work was supported by a grant from the Department of Energy's Office of Science and performed under "Manipulation of Atomic Ordering for Manufacturing Semiconductors" (or µ-ATOMS), a DOE Energy Frontier Research Center. The research team is made up of 10 institutions that since 2022 have collaborated to discover underlying scientific principles determining the order of atoms in semiconductor alloys. 

THE QUANTUM WELL 

Imagine a marble rolling on a groove between two raised edges. It can only move in two directions: back and forth. A quantum well, the thin layer of semiconductor material between two barriers of another semiconductor material with a different energy level, has a similar effect on electrons and holes, or the positive space that an electron can fill. Confined in a well a few atoms thick, the electrons' movement is limited and only certain energy levels are allowed. 

By constraining the movement of electrons and holes, the quantum well makes their motion more predictable and efficient. How easily the electrons or holes move through the well is called its mobility. 

Research on quantum wells began in the 1970s. The stack of material that creates a quantum well must have few defects, and crystal growth is used to produce this nearly pure material. 

Today, quantum wells are widely used to engineer lasers, infrared sensors, more efficient solar cells and high-speed computer chips. 

AN UNEXPECTED RESULT  

Researchers have typically studied germanium-tin quantum wells with barriers of pure germanium. The stack of a germanium-tin well surrounded by silicon-germanium-tin barriers, however, has mainly been examined for optical applications. The silicon-germanium-tin barriers would integrate better with commonly used silicon-based electrical components. 

The U of A produced the critical, high-quality quantum well material for Sandia to build the experimental devices and analyze their electrical performance. Dartmouth analyzed the atomic short-range ordering in the silicon-germanium-tin barriers to help understand the electrical performance. All three institutions collaborated on the experimental design and characterization studies. 

The team of researchers found, to their surprise, that the silicon-germanium-tin barriers created germanium-tin quantum wells with higher mobility. They had assumed that the presence of silicon and tin in the barriers would lead to lower mobility. 

Until recently, scientists were unsure how the trace elements in a semiconductor, such as the silicon and tin in silicon-germanium-tin, are arranged. Are they scattered randomly throughout the main material? Or do they consistently arrange themselves in relation to the main material, a phenomenon known as short range order? 

Early this year, research led by the Lawrence Berkeley National Laboratory and George Washington University found that these trace elements do exhibit short range ordering. Yu and Dartmouth's Jifeng Liu, a co-author on the quantum well paper, contributed to the research on short-range ordering.  

Short-range ordering could explain why the silicon-germanium-tin barriers produced a quantum well with higher mobility. If further research confirms that hypothesis, it could pave the way for manipulating the arrangement of those atoms to dramatically improve performance. 

"It is exciting to reveal the potential impact of atomic short-range ordering on the electrical performance of quantum wells," Liu said. "It offers a new degree of freedom for device engineering." 

"The unexpected high mobility result hints at short-range order effects in the Group-IV SiGeSn system, which is particularly exciting due to the system's optical properties and its potential for monolithic integration with conventional Si CMOS. This short-range order may provide an additional control knob, beyond alloying and strain, for engineering material properties that will impact national priorities in microelectronics and quantum information science," said Sandia's Chris Allemang and first author of the work. 

Yu added, "Even on that tiny scale on the order of a nanometer, you still have hundreds of thousands or millions of atoms. That means you have a larger room to play to enhance the property." 

The latest issue of the journal Advanced Electronic Materials published the research in a paper titled, "High Mobility and Electrostatics in GeSn Quantum Wells with SiGeSn Barriers." The other authors were Christopher R. Allemang, David Lidsky, Peter Sharma and Tzu-Ming Lu of Sandia National Laboratories, Shang Liu of Dartmouth College, and Yusheng Qiu of the University of Arkansas. The paper was featured on the cover of the journal.

Contacts

Fisher Yu, professor
Department of Electrical Engineering and Computer Science
479-575-7265, syu@uark.edu

Todd Price, research communications specialist
University Relations
479-575-4246, toddp@uark.edu